Unable to ping the ESXi host intermittently or host shows 'Not Responding' in vCenter

Products

• VMware vSphere ESXi

Issue/Introduction

The ESXi host intermittently enters a "Not Responding" or "Disconnected" state in vCenter and becomes unpingable. This behavior is typically observed following a host re-installation, firmware upgrade, or a change in the management network configuration (e.g., moving from a Distributed Switch to a Standard Switch).

Symptoms:

• Intermittent ping loss to the ESXi management IP (vmk0).
• vCenter reports the host status as "Not Responding."
• Packet capture (pcap) analysis shows:
  • Repeated TCP retransmissions from vCenter to the ESXi host.
  • High volume of TCP Reset (RST) packets.
  • Broadcast ARP Requests from the ESXi host that do not receive corresponding Unicast ARP Replies.

Environment

ESXi 7.x
ESXi 8.x
ESX 9.x

Cause

The issue is caused by a Teaming Policy Mismatch between the ESXi vSwitch and the upstream physical switch.

1. If the physical switch has an EtherChannel (LAG) configured: The ESXi vSwitch must be set to "Route Based on IP Hash." If left at the default "Route Based on Originating Virtual Port ID," connectivity will be intermittent or lost.
2. If the ESXi vSwitch is set to "Route Based on IP Hash" WITHOUT a physical LAG: The physical switch will be unaware of the link bundle and may route return traffic (ARP replies/TCP ACKs) down an incorrect uplink. ESXi will silently discard this traffic, leading to session timeouts and the "Not Responding" state.

Resolution

Align the teaming policies between the virtual and physical layers by implementing one of the following options:

Option 1: Align ESXi with Physical EtherChannel If the physical switch is already configured with a Static EtherChannel, change the port group or vSwitch balancing to IP Hash.

• Using CLI: esxcli network vswitch standard portgroup policy failover set -p "Management Network" -l iphash
• Using UI:
  1. In the vSphere Client, navigate to the host.
  2. On the Configure tab, expand Networking and select Virtual Switches.
  3. Navigate to the Teaming and Failover policy for the standard switch or port group.
  4. From the Load balancing drop-down menu, select "Route based on IP hash".

Option 2: Align Physical Switch with ESXi (Or Revert to Default) If no EtherChannel is configured on the physical switch, ensure the ESXi vSwitch is set to the default "Route Based on Originating Virtual Port ID". This requires no specialized configuration on the physical switch.

Additional Information

Understanding IP Hash load balancing
Specifying or changing the load balancing policy from the command line in ESXi
Intermittent network connectivity on ESXi or host disconnect issues